Frequency response based system identification and controller tuning
Yousef, Ali Awad Hasan
Master of Science
SubjectAutomatic control Mathematical models
System identification Mathematical models
Feedback control systems Mathematical models
MetadataShow full item record
A detailed simulation based analysis of different frequency response and transfer function identification techniques are performed and critically studied. Standard closed loop and open loop recursive least squares system identification techniques are studied along with methods based on Laguerre filter, frequency sampling filter and bandpass filter. Based on this critical study, the bandpass filter approach is shown to have relatively more advantages than disadvantages over the other techniques of system identification. Methods for PID controller tuning using finite and discrete set of frequency response points are developed for SISO systems and extended to MIMO systems. The controller tuning methods are supported with simulations and experiments on a temperature control system and a distillation column. Based on the results obtained, it is concluded that these methods are practical and capable of being used in industrial process control.